Cancer is one of the leading causes of death worldwide, and cancer, especially lung cancer, is difficult to diagnose and treat effectively. Accordingly, there is a need in the art for new compositions and methods for assessing and treating various cancers, particularly lung cancer.
Lung Cancer
Lung cancer is the second most prevalent type of cancer for both men and women in the United States and is the most common cause of cancer death in both men and women. The five-year survival rate for lung cancer continues to be poor at only about 8-15%. This low survival is because lung cancer is commonly not detected until it has spread beyond the lungs. Only 16% of new lung cancer cases in the United States are detected at the earliest stage, when the cancer is still localized to the lungs. At this early stage, survival is considerably higher, with estimates as high as 70-80%. Therefore, procedures for detecting lung cancer are of critical importance to the outcome of a patient since these procedures have the potential to reduce mortality. Thus, there is a need for new diagnostic compositions and methods that are more sensitive and specific for detecting early lung cancer.
Furthermore, there is also a need for new diagnostic compositions and methods for determining the stage of a patient's disease. Stage determination has potential prognostic value and provides criteria for designing optimal therapy. Biomarkers that are indicative of different stages of lung cancer would be useful to facilitate the staging of lung cancer.
Lung cancer patients are typically monitored following initial therapy and during adjuvant therapy to determine their response to therapy and to detect persistent or recurrent disease or metastasis. Thus, there is clearly a need for lung cancer markers that are more sensitive and specific in detecting lung cancer, its recurrence, and progression.
Although imaging modalities, such as computed tomography (CT) screening, are being studied to aid in the early detection of lung cancer, controversy remains as to the ability of these methods to impact mortality (I-ELCAP Investigators, NEJM 2006 (355):1763-71 and Bach et al. 2007. JAMA 297:953-961). In addition, the most advanced imaging technologies under study are expensive and not widely available. These CT imaging tests may lead to over-diagnosis of lung cancer, resulting in significant expenses to the health care system to manage patients with pulmonary nodules observed through these CT imaging tests. Furthermore, there is significant morbidity associated with the management of the pulmonary nodules in an effort to ascertain whether the nodules are malignant or benign. It is estimated that 10-50% of smokers in a high risk group have pulmonary nodules upon imaging studies (CHEST 2007 Supplement—Evidence for the Treatment of Patients With Pulmonary Nodules When Is It Lung cancer?: AACP Evidence-Based Clinical Practice Guidelines). Thus, there is a significant need for novel diagnostics that can be used either independently or with imaging modalities for early diagnosis and improved management of patients with lung cancer. For example, a blood test for biomarkers that has high performance (e.g., high sensitivity and specificity) for detecting lung cancer could provide a low cost complement to CT testing for early detection of cancer. If the performance of a biomarker test were sufficiently high, such a test could serve as a lower cost alternative to CT or X-ray testing. For example, only those patients that tested positive in a biomarker test may then need to undergo more expensive imaging tests. Furthermore, a biomarker test could be used, for example, in a yearly screening regimen for lung cancer.
Although there have been reports of circulating tumor markers and antigens with potential use in lung cancer (see Schneider, J. 2006. Advances in Clin Chem, 42: 1-41 for a review), markers currently used generally suffer from low sensitivity and less than desirable specificity, especially among smokers (Schneider, 2006), and are typically only used to monitor for recurrence of lung cancer. Thus, there is a need in the art for a panel of markers with high sensitivity (and varying specificities, depending on the clinical indication), such as for detecting lung cancer. Furthermore, there is also a need for novel markers that are useful individually or as part of a panel for detecting lung cancer. Such markers, and panels of markers, would facilitate management of patients with lung cancer, for example.
For a further review of lung cancer diagnostics, including the use of tumor biomarkers as well as CT screening, see the following citations: Schneider, “Tumor markers in detection of lung cancer”, Adv Clin Chem. 2006; 42:1-41; Bach et al., “Computed tomography screening and lung cancer outcomes”, JAMA. 2007 Mar. 7; 297(9):953-61; and International Early Lung Cancer Action Program Investigators et al., “Survival of patients with stage I lung cancer detected on CT screening”, N Engl J. Med. 2006 Oct. 26; 355(17):1763-71. Also see Pepe et al., “Phases of biomarker development for early detection of cancer”, J Nat'l Cancer Inst. 2001. 93(14):1054-1061
Description of Tables 1-2
Tables 1 and 2 provide further information for lung cancer markers (“LCM”), including their names, symbols (alternative symbols are indicated in parentheses), Genbank protein accession numbers, and an exemplary protein sequence for each marker (except for the carbohydrate antigens CA 242, CA 19-9, and CA 72-4, for which representative journal citations are provided for each). Exemplary LCM protein sequences are provided as SEQ ID NOS:1-65 (additionally, the carbohydrate antigens CA 242, CA 19-9, and CA 72-4 are also provided). Nucleic acid sequences (e.g., mRNA transcript sequences and genomic DNA) and alternative protein sequences for each marker are well known in the art and can readily be derived using the information provided in Tables 1-2, for example.
The LCM provided in Table 1 are as follows (alternative names/symbols are indicated in parentheses): SLPI, MIF, TIMP1, TFPI, ENO2 (NSE), CEA (CEACAM5), MMP2, AMBP, Cyfra 21-1 (Cyfra, KRT19), SCC (SERPINB3), OPN, defensin (DEFA1, HNP-1, HNP1-3), CA 242, CA 19-9, CA 72-4, MN/CAIX (CA9), ProGRP (GRP), KRT18 (TPS), ECAD (CDH1), TIMP2, CD44, LGALS3BP, ERBB2 (HER-2), UPA (PLAU), DKK (DKK1), CHGA, VEGF, KITLG, PBEF (visfatin), SORT1 (sortilin), MDK (midkine), IGFBP3, IGFBP4, CTSC, ICAM3, CTGF, LCN2, EGFR, BGN, TIMP3, HGF, MUC16 (CA125), NCAM, CRP, SERPINA1 (ATT), PKM2, RBP, KLK11, KLK13, SAA, and APOC3.
The LCM provided in Table 2 (which are particularly useful as autoantibody markers) are as follows (alternative names/symbols are indicated in parentheses): TP53 (p53), KLKB1, CFL1 (CFLN), EEF1G, HSP90α (HSP90AA1), RTN4, ALDOA, GLG1, PTK7, EFEMP1, SLC3A2 (CD98), CHGB, CEACAM1, ALCAM, HSPB1 (HSP27), LGALS1, and B7H3.
Elevated levels of each of these LCM are indicative of lung cancer, except for sortilin (SORT1), for which low levels are indicative of lung cancer.
Description of Tables 3-12
Table 3 provides 35 different panels of 11 markers (each row of 11 markers represents a panel) that have at least 98% specificity and 82% sensitivity for detecting lung cancer. The total number of occurrences of each marker in these 35 11-marker panels is indicated at the bottom of Table 3. Seven markers (SLPI, TIMP1, TFPI, SCC, OPN, CEA and CA242) appear in all 35 of these panels, GRP appears in 33 of these 35 panels, MIF appears in 29 of these 35 panels, and NSE and HNP-1 each appear in 15 of these 35 panels. AMBP, Cyfra, MMP2, Ca72-4, Ca19-9, and CAIX each appear in 7-9 of these panels, as indicated in Table 3.
Table 4 provides markers the can be included in any of the panels disclosed herein. For example, the markers in Table 4 can be added to any of the panels disclosed herein and/or can replace one or more members of any of the panels disclosed herein. As a specific example, the markers in Table 4 can be added to any of the panels disclosed in Table 5 and/or can replace one or more members of any of the panels disclosed in Table 5. The markers disclosed in Table 4 are also disclosed in Table 2.
Tables 5-12 provide data for the analysis of various panels in various lung cancer uses, such as distinguishing lung cancer samples versus normal samples such as for diagnosing/detecting lung cancer (Tables 5-6 and 11-12, for example), as well as certain specific uses (these specific uses, which may be referred to herein as “indications” or as determining or assessing lung cancer “characteristics”, are provided in Tables 7-10, for example). In Tables 5-12, each row represents a panel (a panel may comprise an individual marker). For each panel in Tables 5-12, data are presented based on logistic regression and/or split point analysis (as indicated in each table). Area under the curve (AUC), sensitivity at 95% specificity, and specificity at 95% sensitivity are indicated. “Size” (second column) indicates the number of markers in the given panel. Further information regarding characteristics of the sample sets (the “54×53”, “50×50”, and “44×44” sample sets) used in each of the analyses is provided in FIG. 16 (the “104×103” sample set used in Table 8 is the “54×53” and “50×50” sample sets combined). In Tables 5-12, and elsewhere herein, “trained” refers to the sample set (which may be referred to as the “training set”) which was used to formulate cutoff levels, and “tested” refers to the sample set (which may be referred to as the “testing set”) to which these cutoff levels were applied (such as to classify a sample as a lung tumor or normal sample, or other specific use, based on whether marker levels were above or below the cutoff levels established from the training set).
Table 5 provides data for logistic regression and split-point analysis of the 9-marker panel of Cyfra, SLPI, TIMP1, SCC, TFPI, CEACAM5, MMP2, OPN, and MDK, and all subcombinations thereof (including individual markers), in distinguishing lung tumor samples versus normal (i.e., control/healthy) samples, such as for diagnosing/detecting lung cancer. For each panel in Table 5, data are presented based on logistic regression and split point analysis and based on analysis of either training and testing on the same 54×53 (54 controls and 53 cases) sample set, or training on the 54×53 sample set and testing on the 50×50 (50 controls×50 cases) sample set (see FIG. 16 for characteristics of these sample sets). Area under the curve (AUC), sensitivity at 95% specificity, and specificity at 95% sensitivity are indicated. The panels are sorted based on the AUC indicated in the third column. “Size” (second column) indicates the number of markers in the given panel. Thus, Table 5 provides the 9-marker panel of Cyfra, SLPI, TIMP1, SCC, TFPI, CEACAM5, MMP2, OPN, and MDK, and all panel subcombinations thereof, including each of these nine markers individually (each row represents a panel).
Table 6 provides data for split-point analysis of panels (including individual markers) that include any of the nine markers in the panels provided in Table 5 and/or various other markers (which are not in the panels provided in Table 5) in distinguishing lung tumor samples versus normal (i.e., control/healthy) samples, such as for diagnosing/detecting lung cancer.
Table 7 provides data for logistic regression analysis of the 9-marker panel of Cyfra, SLPI, TIMP1, SCC, TFPI, CEACAM5, MMP2, OPN, and MDK, and subcombinations thereof (including individual markers), in distinguishing adenocarcinoma versus squamous cell carcinoma types of lung cancer.
Table 8 provides data for split-point analysis of the 9-marker panel of Cyfra, SLPI, TIMP1, SCC, TFPI, CEACAM5, MMP2, OPN, and MDK, and subcombinations thereof (including individual markers), in distinguishing stage I versus stage III lung cancer. In addition to their utility in distinguishing between early and late stage lung cancer (e.g., stage I or II versus stage III or IV), the panels provided in Table 8 are also useful for distinguishing between any other stages of lung cancer (e.g., any of stages I, II, III, and IV).
Table 9 provides data for split-point analysis of various panels in distinguishing small cell lung cancer (SCLC) versus other types of lung cancer (e.g., non-small cell lung cancer, NSCLC). In the left-side of Table 9, marker levels are higher in NSCLC (as compared to SCLC). In the right-side of Table 9, marker levels are higher in SCLC (as compared to NSCLC).
Table 10 provides data for split-point analysis of the 9-marker panel of Cyfra, SLPI, TIMP1, SCC, TFPI, CEACAM5, MMP2, OPN, and MDK, and subcombinations thereof (including individual markers), in distinguishing malignant lung tumors versus benign lung lesions.
Table 11 provides data for split-point analysis of various panels in distinguishing small cell lung cancer (SCLC) versus normal (i.e., control/healthy) samples.
Table 12 provides data for split-point analysis of various panels in distinguishing lung cancer (including both small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC)) versus normal (i.e., control/healthy) samples.